Development of usage statistics for RepositóriUM

Abstract

The usage statistics software developed at Minho University, for RepositóriUM, the institutional repository based on DSpace is presented. The system was developed based on the principles and some of the components of the statistics application created and released by ANU. Despite the fact that the development was made to respond to the specific needs of RepositóriUM, the system is completely adjustable to other environments, because their components can be easily configured, changed or extended, to respond to different information needs. The system is composed by 3 main components: Log Processor, Data Model, Web Interface. The first one is responsible for the gathering, real time processing and storing of event logs generated by DSpace. It is composed by different event processors, each one adjusted to a specific event. The second component is the data model stored in a data base management system (DBMS), where the relevant, and properly processed, information is archived. The Web interface is flexible, allowing to configure – according with a base structure –the display of the information in the user interface. The structure of the system is based in the organization of information in two dimensions: type of information (access, content, administrative) and aggregation level (repository, community, collection, item). These two dimensions are also the base for the definition of user access policies, enabling the definition of different usage profiles. Statistics are configured and grouped in the system as SQL queries. Their presentation (layout) is constructed with a XML-XSL mechanism that can be configured at the individual query level. Other interesting features of the system are: automatic exclusion of crawlers, detection and processing of the origin (country) of accesses, access permissions at data level (e.g., restrict access to data from one particular community to members of that community).